Bicycle Quarterly rolling resistance tests: Spring 2013

[MikeDee]

I mentioned above that Heine had switched to either constant speed or constant power runs -- I think it's the Marymoor velodrome (but I could be wrong) --

Looks like that was correct. Heine has come out with another post on his rolling resistance findings, with some added information about his test protocols. You can read some (vigorous) discussion in the comment thread here:
https://janheine.wordpress.com/2015/01/ ... evolution/

A few years ago I tried suggesting to Heine that there are better ways to estimate Crr (and CdA) in field tests but, um, he has chosen to proceed with his own methods. He probably figured I didn't know what I was talking about which, now that I think about it, I can't blame him for since that is often not bad figuring.

Looks like another nail in the coffin for stiff walled tubeless tires.

[climbandpunishment]

Pardon me if it's already been mentioned, but keep in mind the aerodynamic differences between tires quite probably vary quite a bit based on what wheel they're tested on, due to the way the different tire textures will affect the boundary layer.

Also, I really appreciate the technical level of discussion in this thread. I don't have the expertise to evaluate the validity of all the different tire resistance tests, so thanks for all your guys' really in-depth thoughts and effort on this.

[TheKaiser]

Pardon me if it's already been mentioned, but keep in mind the aerodynamic differences between tires quite probably vary quite a bit based on what wheel they're tested on, due to the way the different tire textures will affect the boundary layer.

I haven't seen the texture specifically referenced, other than Zipp having dimpled tires in the past (although they seem to have stopped that on newer models). Several people have discussed how GP4000s seem to offer superior aerodynamics on most rims though, due to their more oval shaped cross section which helps smooth the transition of airflow around the casing and onto the rim, vs. a rounder cross sectioned tire. Also, open tubulars are known to have worse aero properties, all else being equal, due to the abrupt transition from the tread cap to the casing, which creates a lip, whereas vulcanized tires are more uniform and smooth in this area.

I would be curious to see more info on effects of the tread though if it exists.

[TheKaiser]

One more thing regarding tire tread impacting the aero properties, which then can confound open air RR tests...have any of you ever pedaled a fully treaded MTB up to speed on a repair stand? You can feel the air getting flung off the tire by the knobs, almost like it is a really inefficient fan. I have pondered the mass of this air, and what the power consumption of its displacement is, but never seen any concrete figures. At some point, the boundary layer benefits of minimal tread may outweigh the fan effect I just mentioned, but where that point is I have no idea.

It is a bit like how when riding in the rain your tire picks up water and flings it up in a rooster tail, which has to suck some power, but at the same time, the wetness of the pavement will potentially reduce the friction vs. dry conditions. Some minimal amount of wetness is beneficial, but there has to be a crossover point where the water starts adding more resistance than it reduces, however the calculation of the two is very difficult.

[Marin]

It might be extremely hard to calculate, but I'm sure one can design a setup to test both cases without too many problems!

[climbandpunishment]

Kaiser, that's really interesting; I hadn't considered that effect. It's actually a fairly well-studied flow, since the flow is set up almost entirely due to the viscous entrainment from the rotation:



If you're curious and/or bored, you can look for "viscous flow rotating disc" for more info. Frank White's book Viscous Flow has a good description of it from what I remember.

It'd be feasible to determine for a disc, but a real-world wheel might be another story, since you'd have to estimate the viscous contribution from the spokes as compared to that of the rim and tire. My SWAG would be that it'd be a secondary effect since it's primarily viscous in nature. I would tend to think that different (road) tires would be pretty similar in terms of this rotation effect. Mountain tires, which are both larger in area relative to the wheel and bumpier, might be a different story.

And getting back to the likely bigger effect of the tire tread, that's cool about tubulars/GP 4000s. Even a tiny disturbance at bike speeds/sizes would have a significant effect when it comes to texture. For a recent bike frame design, I calculated that the bump/texture size needed to trip the flow from laminar to turbulent (and potentially keep the flow from becoming separated, i.e. messy and high drag) for a tire+rim is only about 0.03-0.09 mm, depending on rim depth and speed. So even road crud could potentially have a significant effect, much less actual tire texture.

[TheKaiser]

For a recent bike frame design, I calculated that the bump/texture size needed to trip the flow from laminar to turbulent (and potentially keep the flow from becoming separated, i.e. messy and high drag) for a tire+rim is only about 0.03-0.09 mm, depending on rim depth and speed. So even road crud could potentially have a significant effect, much less actual tire texture.

That is really cool to see some actual numbers put to these things! So I guess that maybe that is why Zipp ditched the dimples in favor of a more minimal and traditional siped tread. If something as small as road grit is sufficient to provide aero benefit, then you are better off designing a tire for better RR and traction.

On an unrelated but relevant tangent, I have also pondered the effect of treads on RR figures. Conventional wisdom suggested that they had no effect on traction, but more recent testing (by Wheel Energy in Finland if memory serves) suggests that the fine mechanical linking of tread with the road surface actually does enhance traction. That suggests to me that there is also the potential for increased frictional losses from road surface contact, although on further thought, if the file tread allows for thinner rubber between the tread bumps that could lower hysteresis vs. a tire with similar overall rubber thickness but a uniform depth and no tread.

[climbandpunishment]

That is really cool to see some actual numbers put to these things! So I guess that maybe that is why Zipp ditched the dimples in favor of a more minimal and traditional siped tread. If something as small as road grit is sufficient to provide aero benefit, then you are better off designing a tire for better RR and traction.

That'd be my guess. I forgot to mention, but another key point is that you really only want texture of that very small 0.03-0.09 mm size to trip the flow. Once you've got texture large enough to trip the flow, any larger and you're just adding drag. As an aside because I'm a nerd, Zipp's wheel dimples are larger because they're meant to cause separation in a controlled manner rather than tripping flow and preventing it - they have to modify the flow on a larger scale rather than simply tripping the very thin boundary layer of air right near the wheel surface, so they are necessarily larger. You need a very well-controlled surface to do that, and (wear-prone) tires aren't good for that.

Anyways, Zipp's designs are smart enough that I'd need to do some pretty hefty CFD (i.e., low end supercomputer-level stuff) to figure out exactly what their aerodynamic strategy for the wheel/tire system as a whole really is, but my guess is that you're correct.

And that sounds like a cool finding from Finland. I'm definitely no tribologist, but I wonder whether you could tailor tread sizes to road surface roughness for better grip.

[jooo]

Also, open tubulars are known to have worse aero properties, all else being equal, due to the abrupt transition from the tread cap to the casing, which creates a lip, whereas vulcanized tires are more uniform and smooth in this area.

I'm not sure where, but I'm sure I saw it written somewhere that part of the gain was also due to a smoother transition from rim to tyre that can be had with some combos. Almost as if the they'd were a single piece, rather than having quite a lip like some tubular rim/tyres can end up with - or the thing Mavic blades are meant to cover up.

edit: whoops, didn't realise you were talking about 'open' tubulars.

[TheKaiser]

Not sure if this is old news to you guys, but I stumbled across a new (to me) source for RR data. I was already aware of Al Morrison, Wheel Energy, Tour Mag, BQ, etc..., but this is a guy in the Netherlands doing his own 77cm textured steel drum tests. It looks like he has had this website since the middle of 2014, and he had answers to a few questions that I have pondered, like what effect does adding sealant have on RR in a tubeless tire (apparently not very much, but it is dependent on the volume added).

From the little I have seen of his data, most of his figures fit in pretty well with the other sources, so it seems pretty consistent with known good data: http://www.bicyclerollingresistance.com/

[istigatrice]

That's an interesting source, but the steel roller will heavily bias the conti tyres, a stiff tyre is not penalised as much as we expect it would be on the road. They also haven't validated the effects of the "diamond" pattern - they're using it as a model, and any model should be validated (or at least not rejected) with real world conditions before it can be accepted. Hopefully they'll get around to doing that or at least making a note of this point.

OT: Also from a scientific point of view, they claim that their tests are very ACCURATE, stating that they only vary by 0.3 watts between runs. This is nonsense. This is an example of precision, not accuracy. They've made a great precise test, but we have no idea about his accuracy, and they claim to have great accuracy, yet haven't substantiated why. I'm sceptical because accuracy doesn't matter if we're only looking at trends. They should clearly substantiate this point so that they can demonstrate they understand what they're doing. Right now to me they seem like a no-clue who's spitting out numbers without context or relevance.

They present some interesting trends, which could be useful - they need to substantiate why certain tyres have been reshuffled in the Crr order (eg rubino pro faster than Pro4) and why higher pressures result in lower Crr (evidence for their drum biasing stiffer tyres).

[droid1973]

IMHO I think that the go wider fad is purely marketing

[dogg]

well thats just like, your opinion man.

[Zoro]

This is an example of precision, not accuracy. They've made a great precise test, but we have no idea about his accuracy, and they claim to have great accuracy, yet haven't substantiated why. ...

Exactly right.
I've been using tubular tyres since late 1970s. My wife, son and I have a few hundred thousand K (all train, and he races mostly only on tubulars).
I concluded lab tests are not that helpful to tell what real world use will be like.

My experience is how a tyre behaves on rollers is very different than the road - and the rollers.
We use for race warm-up rollers with a 5" drum, 3" drum (mostly), 2.5 (the small ones) with a power meter and heart rate monitor on the rider.
There is not an obvious correlation between tyres switching from a 5" to the 2.5" (so best on 5" may not be best on 2.5") and each combo requires different pressure to get the same rider HR/power resistance - at steady speed. Change the speed and one combo increases resistance much faster than another tyre.

When Pro Tour teams use a tyre they buy and block out the brand name, you have a good idea what the best tyres are - all around.
You can hold and feel a tyre and guess a lot about it depending on stiffness of case etc. A real test in the real use case is the only real useful test. That and factor in durability issues based on road and rider habits.

The Tufos get bad rolling reviews and I personally don't like them, but they are not designed to be ridden at 100PSI, the FMB silks and Vittoria's are. So each tyre should be tested at its sweet spot pressure, speed, deflection (or total downward force and road bumps), and have tyre weight, road condition, puncture resistance and other factors guessed in to get a close result.

[kgt]

Nice posts by istigatrice and Zoro.